Uncategorized

AI Search and Funnel Stage: What a Digiday Survey of 142 Marketers Reports

Many marketers suspect AI search is reducing their traffic, but few measure it by funnel stage. A Digiday survey of 142 professionals offers one data point on where the losses are showing up and how teams are responding. This article reads it carefully, including its limits.

Key Takeaways
  • In Digiday’s 2026 survey, 37% of respondents reported upper-funnel search traffic decreases due to AI, against 21% for lower-funnel traffic.
  • The most common GEO response, at 34%, was adding more content that answers consumer questions.
  • 16% said they were not familiar with GEO or AEO, so awareness is still uneven.
  • The sample is small and self-reported, so use it to shape your own measurement, not as a benchmark.

What the survey covered

GEO: Generative Engine Optimization, the practice of making content likely to be cited in AI-generated answers. AEO: Answer Engine Optimization, a closely related term focused on being the direct answer.

Digiday published its research on May 11, 2026. It surveyed 142 brand and agency professionals drawn from Digiday’s own audience and added interviews with marketing and technology executives. Field dates were not stated in the article. Two limits follow. The respondents are self-selected readers of a marketing trade publication, and the answers are perceptions, not analytics data. A result such as “37% saw decreases” means 37% of this group said so. It does not estimate how many businesses overall lost traffic.

Upper funnel versus lower funnel

Digiday reported that 37% of respondents said their companies saw upper-funnel search traffic decrease because of AI, while 32% said it was unchanged. For lower-funnel traffic, 21% reported decreases and 36% reported no change. The article notes that the figures do not account for every other response option.

Search traffic typeReported decreaseReported unchanged
Upper-funnel37%32%
Lower-funnel21%36%

The pattern is plausible, since informational queries are the easiest for an AI summary to answer. But it is still a perception gap in a small sample. The implication for you is to check your own data by query type before changing budget, because a loss in broad informational traffic may matter less than a loss in purchase-intent queries.

How teams are responding

Digiday asked what respondents were doing about GEO and AEO. The most common answer, from 34%, was highlighting more content that answers potential consumer questions. Another 27% said they had not changed strategy, and 16% said they were not familiar with GEO or AEO.

Those numbers describe a market still deciding. More than four in ten respondents either had not changed anything or did not know the terms. For a business evaluating outside help, this suggests that vendors who claim a settled playbook are overstating the state of practice. Our guide to what to put in a GEO engagement scope covers how to test those claims. The leading tactic, answer-focused content, is also modest and low risk, which is why it is a reasonable starting point.

What to measure in your own data

Because the survey cannot tell you about your site, build your own baseline. Follow these steps in order.

  1. Group your Search Console queries into informational and commercial intent, and compare clicks year over year for each group.
  2. Segment landing pages the same way, so top-of-funnel guides are separate from product and service pages.
  3. Track AI referral visits separately, using the approach in our guide to tracking AI referral traffic.
  4. Compare conversions, not only sessions, so you can see whether lost clicks were ever valuable.
  5. Record changes you make, such as new answer-focused pages, so results can be tied to actions.

This turns a vague worry into a testable claim. It also fits the sizing logic in our piece on how much to invest in AI search.

Frequently Asked Questions

Does the survey prove AI search is reducing traffic?

No. It reports what 142 professionals said about their own companies, not measured traffic, so it indicates sentiment rather than proof.

Why would upper-funnel traffic fall faster?

Broad informational queries are the kind AI summaries can answer directly, though the survey itself does not test that explanation.

What is a sensible first GEO step?

Add clear, sourced answers to the questions your buyers ask, then measure query and conversion changes against a baseline.

Reading survey data critically is part of deciding where to invest. At Grow with Hafsa we help US and UK businesses build funnel-level baselines for AI search and set sensible scopes, as part of our AI marketing and performance marketing services.